Hand foot and mouth disease (HFMD), which was first reported in New Zealand in 1957, is an unremarkable illness that commonly occurs in young children, most of whom are younger than 5 years old. Two closely related viruses, coxsackievirus A16 (CA16) and enterovirus 71 (EV71), which belong to the Enterovirus genus of the Picornavirus family, and usually co-circulate during HFMD outbreaks ^[1]–^[3], have been identified as the most frequent pathogens of HFMD, and other enteroviruses, including CA2, CA4, CA5, CA7, CA9, CA10, CA16, CB1, CB2, CB3, CB4 and CB5 can also cause HFMD. In the most common manifestation which gives the syndrome its name, children typically present with vesicular exanthema on the soles of their feet, the palms of their hands and in their mouths, causing discomfort and feeding difficulties. However, severe health consequences or deaths may occur owing to complications such as encephalitis, aseptic meningitis, acute flaccid paralysis, myocarditis, autonomic nervous system dysregulation (ANS) and pulmonary edema (PE) or hemorrhage.

In recent years, outbreaks of HFMD have increased and more and more severe cases have appeared, especially in China ^[4], which has posed a persistent and severe public health problem. Severe HFMD cases have many characteristics in laboratory diagnosis. Hyperglycemia and leukocytosis are two important characteristics in severe cases of HFMD, and blood glucose and white blood cell (WBC) counts change with the severity of the disease, which have been addressed by multiple studies such as those reported by Mao et al ^[5] and Yu et al ^[6], and the results are not always consistent. Meta-analysis and systematic review can solve these disputes. To our knowledge, these data have not been systematically evaluated and reported in English. The objective of this study was to summarize the data and evaluate the characteristics of these indexes in patients of HFMD using a meta-analysis of published articles.

As shown in Fig. 1, a total of 422 articles published in Chinese or English were identified. After excluding 4 review papers and 377 original articles without data about healthy controls and mild cases of HFMD or mild cases and severe cases of HFMD or CNS and ANS/PE or cases infected by EV71 and cases infected by CA16 based on abstract evaluation, 45 studies were selected and the full-text versions were retrieved. Of these, 8 were excluded based on the exclusion criteria. Finally, 37 studies ^[5], ^[6], ^[12]–^[48] were included in this review. Among them, 33 studies, from Mainland China, and 4 studies, from Taiwan, addressing blood glucose and/or WBC counts in healthy controls and mild cases of HFMD, or in mild cases of HFMD and severe cases of HFMD, or in CNS and ANS/PE, or in cases infected by EV71 and cases infected by CA16 were presented in Table S1, Table S2, Table S3 and Table S4.

Repeated outbreaks of HFMD have occurred in many parts of the world such as Malaysia ^[49]–^[50], Taiwan ^[1], ^[51]–^[54], Singapore ^[55], Mainland China ^[56], Brunei ^[57], Western Australia ^[58], the Unites States ^[59] and Germany ^[60]. The first reported large and severe HFMD epidemic occurred in Taiwan region in 1998 including about 130 000 cases of HFMD, among whom 405 patients were severe cases and 78 died ^[1], ^[51]–^[52]. In 2000, there was another report of outbreak, with 80 677 cases of HFMD and 41 deaths there ^[52]. In Mainland China, HFMD has occurred repeatedly in recent years and more and more cases have been reported. There were 83 344, 488 955 and 1 155 525 cases in the nationwide in the year of 2007, 2008 and 2009, respectively, reported by the Ministry of Health of the People's Republic of China. The corresponding deaths increased with years and were 17, 126 and 353 for these years, respectively. This suggested that HFMD had become a severe public health problem in China. Therefore it is necessary to find feasible indexes to monitor the severity and prognosis of HFMD patients.

When children suffer from HFMD, the prevalence of hyperglycemia, the prevalence of leukocytosis, the level of blood glucose and WBC counts will change with the severity of the disease. In this article, we performed a meta-analysis on these indexes respectively. To the best of our knowledge, this is the first time to apply meta-analysis to the prevalence of hyperglycemia, the prevalence of leukocytosis, the level of blood glucose and WBC counts in cases of HFMD.

By using a search strategy, and inclusion criteria and exclusion criteria, 37 studies were selected including 33 studies reported in Mainland China and 4 studies reported in Taiwan. Far less was known regarding the descriptive epidemiology of HFMD in countries outside the Western Pacific Region ^[4]. It is unfortunate that no studies reported in countries other than Mailand China and Taiwan fulfilled our inclusion criteria to be included in our analysis. Therefore our research focused on the characteristics of blood glucose and WBC counts reported in Mainland China and Taiwan. When children suffer from HFMD, the patients could present different symptoms from the mild to the severe such as HFMD, encephalitis, ANS and PE. The characteristics of blood glucose and WBC counts which have been previously reported are inconsistent. For example, one study ^[6] claimed that the average level of blood glucose in mild cases of HFMD was higher than that in healthy controls, but another study ^[5] claimed that there was no significant difference between them. As another example, some studies ^[8], ^[13], ^[27], ^[37] claimed that WBC counts in ANS/PE were higher than those in CNS but other studies ^[24], ^[38] showed that there was no significant difference between them. By meta-analysis we solved the divergence between them and concluded that the level of blood glucose, the prevalence of hyperglycemia, WBC counts and the prevalence of leukocytosis increased with the severity of the disease, but there was no significant difference in the average level of blood glucose between mild cases of HFMD and healthy controls. In the symptom-matched cases infected by EV71 and CA16, respectively, WBC counts were less in cases infected by EV71 than those in cases infected by CA16. However, there were not enough data available for meta-analysis of the level of blood glucose between cases infected by EV71 and CA16, and of WBC counts between mild cases of HFMD and healthy controls.

Blood glucose and WBC counts could help doctors manage patients in 2 ways by judging the severity and prognosis. First, blood glucose and WBC counts could reflect the severity of the illness. Lin et al ^[9] reported that both the level of blood glucose and WBC counts increased greatly from the mildest uncomplicated group to the severest encephalitis with pulmonary edema group when HFMD patients were divided into 3 groups: the uncomplicated group, the encephalitis group and the encephalitis with pulmonary edema group. There were also other such reports ^[10], ^[37] which suggested similar findings. Elevated level of blood glucose could also forecast disease severity. When the level of blood glucose of mild cases of HFMD increased rapidly, these mild cases would develop into severe cases in 2–4 hours ^[13]. Therefore doctors would have enough time to take measures in advance to manage the patients. One report denoted that when the level of blood glucose was controlled under 7.8 mmol/L the survival rate of the patients increased 5.52 times compared to that of the patients with the level of blood glucose >11.1 mmol/L ^[61]. Second, blood glucose and WBC counts could also be used to predict the prognosis of the illness. Elevated blood glucose and WBC counts often indicated poor prognosis. Chang et al ^[19] reported that all the 11 patients with blood glucose levels from 14.1 mmol/L to 42.2 mmol/L and WBC counts from 11.6×10⁹ cells/L to 40.6×10⁹ cells/L died. A patient even died 3 hours later after admission. In another report, Lin et al ^[24] reported that 15 patients with an average level of blood glucose of 7.66 mmol/L and WBC counts of 11.72×10⁹ cells/L all recovered after treatment while 9 patients with an average level of blood glucose of 13.84 mmol/L and WBC counts of 16.36×10⁹ cells/L all had sequelae or died. Therefore blood glucose and WBC counts were very helpful for doctors to manage HFMD patients especially in primary care settings and resource-limited settings such as most of the counties in China. These 2 indexes would also help doctors to manage HFMD patients in other countries of the world. However, they should be used combined with clinical symptoms and other detection methods such as chest x-ray, brain MRI, abnormal heart rate and Troponin I ^[62]–^[63].

The limitations of this study should be kept in mind. First, most of the included studies were published in Chinese, quality of the reports may be variable. Second, in the included articles, the detailed information of the cases of HFMD (e.g. encephalitis, aseptic meningitis, acute flaccid paralysis, myocarditis, pulmonary edema or hemorrhage etc.) could not be obtained; relevant stratified analyses could not be performed to disclose more detailed characteristics of blood glucose and WBC counts in cases of HFMD and it was difficult to analyze the high level of heterogeneity between studies. Third, not all necessary information, such as age and gender of the study population, could be obtained from all included studies.

In conclusion, our analyses suggested that blood glucose and WBC counts increased with the severity of the illness. There was no difference in the level of blood glucose between healthy controls and mild cases of HFMD. WBC counts in cases infected by EV71 were less than those in cases infected by CA16. Our results advise monitoring of blood glucose and WBC counts in patients of HFMD, which would help doctors to manage HFMD patients greatly.

Competing Interests: The authors have declared that no competing interests exist.

Funding: These authors have no support or funding to report.

We thank the following people from our lab: Linqing Zhao, Liying Liu,Yu Sun and Fang Wang, who kindly assisted us and provided comments on the English in the manuscript, which greatly improved the paper.